ASQ Research Discussion - Volume 1, Number 3
ASQ Research Discussion Listserver
Volume 1, Number 3, September 7, 1995
Current subscribers: 283
General listserver guidelines:
- Postings and subscription requests: ASQ@UMICH.EDU.
- Please reply to the moderators, rather than to all subscribers.
In this issue
- Remarks from the moderators.
- Responsese to business age and size question
- a. Rhonda Reger <RREGER@bmgtmail.umd.edu>
- b. William Kaghan <kaghan@u.washington.edu>
- c. Andrew E Newman <anewman@magnus.acs.ohio-state.edu>
1. Remarks from the moderators.
Thank you for your participation in the listserver. We welcome your comments about how the posting mechanism is working for you. We would especially welcome ideas about how to fine tune the procedures in order to make participating in the ASQ listserver as enjoyable as possible for you.
2. Responsese to business age and size question
a. Rhonda Reger <RREGER@bmgtmail.umd.edu>
Please forgive my simpleminded (yet quite sincere) question: Why is age an interesting variable to study? My background is in strategic management, and I know that many studies have linked age of firm with performance, but I don't see how age is an actionable variable. Perhaps this is my business school/strategic management/practical bent bias. Is age a convenient proxy for other variables, or is it of interest directly?
Rhonda K. Reger
Faculty of Management and Organization
College of Business and Management
University of Maryland at College Park
College Park, MD 20742-1815
Phone: 301-405-2167
Fax: 301-314-9157
e-mail: rreger@bmgtmail.umd.edu
b. William Kaghan <kaghan@u.washington.edu>
Will Mitchell's question and the responses put me in mind of some more general questions that I had about the general method of Organizational Ecology. I also have not had a chance to read the two papers but I made a (perhaps big) guess that both relied on some form of modified or controlled form of event history analysis on two different populations. If I am wrong in this, please correct me.
I am not an not a practitioner of event history analysis so perhaps my questions stem from lack of familiarity with the general method. But I have been interested in Org Ecology and try to keep up with the general results. Two things that I consider more theoretical/analytical have always puzzled me. In one of the early survey articles on this literature, I believe Carroll divided the basic approach into three levels: developmental (?), population, and community (?). If I recall correctly, he did not draw clear linkages between the three levels. In one of the later books, I think that Hannan, Freeman, or Carroll or some combination asserted that the population level studies had validity or power independent of any particular micro level studies. I am not sure whether they made similar assertions about population level studies and community level studies.
The original question and the responses seem to suggest that it might be worth taking a look at these issues. Many of the responses seemed to suggest that the community within which the two different populations were embedded might account for the different results. A couple of responses seemed to suggest that developmental level effects were important in accounting for the differing results. I do a lot of field research myself and so I have always wondered whether population level effects were truly independent of developmental level activity.
As I said, I haven't really followed the Org Ecology literature closely so these issues may already have been addressed. But I would be interested in what people have to say on these somewhat more general issues.
c. Andrew E Newman <anewman@magnus.acs.ohio-state.edu>
Mitchell is, indeed, correct that much of the lack of agreement on the relationship between age, size (and other covariates) with organizational survival is due to inconsistency in the coding of organizational end states. This lack of agreement inhibits generalizability and comparison of research results. Often, authors do not bother to explain how they coded their dependent variables. There is, I think, an analogous problem in the study of organizational foundings, which often fails to distinguish between new startups, spinoffs, and other initiating states.
Beyond this, there is (after all these years of research) a striking lack of agreement about the THEORETICAL MEANING of such central operational variables as organizational age and size. (In fact, the appropriate measurement of age and size is more problematic than most researchers suppose, as instantiated by recent work by Amburgey & Barnett, and the cited paper by Barron, West & Hannan). Besides the multiple possible meanings and causal mechanisms underlying age, there is and likely will remain uncertainty whether time-related effects are genuine age effects, or artifacts of unspecified population heterogeneity. My remarks here have less to do with age than with the meaning of size as a source of heterogeneity.
In work recently reported at the meeting of the American Sociological Association in Washington, examining patterns of several forms of change in government organization, I found nonmonotonic age effects on organizational mortality. I also found that inclusion of time-varying organizational size (measured as constant dollar organizational budgets) did not substantially alter estimates of age effects. However, I also found that inclusion of measures of organizational power (essentially, market dominance) vitiated the effects of organizational size (which had been consistent with those found by Barron, West, and Hannan) to the point of statistical nonsignificance. The measures of organizational power were themselves unaffected by the addition of size.
